10 Bits: the Data News Hotlist
This week’s list of data news highlights covers January 30 – February 5, 2016 and includes articles about college freshmen using fitness trackers to fulfill fitness requirements and plans to launch a network of micro-satellites into space.
The Sunlight Foundation, a nonprofit government transparency organization, has launched Hall of Justice, a database of publicly available criminal justice data and research from all 50 states, the District of Columbia, and the federal government. Though some states have made certain criminal justice data sets easily accessible online, much of this data is difficult to access and stored in disparate locations, if it is available at all, and required the Sunlight Foundation to manually compile this data. Sunlight Foundation created Hall of Justice to make it easier for members of the public to analyze criminal justice data and to highlight the need for more consistent and accessible publication of this data throughout the United States.
Oral Roberts University, which includes regular physical activity as a college requirement, now requires students to wear the Fitbit wearable fitness tracker to streamline the old complicated system of manually recording physical activities to meet weekly quotas. Now, all freshmen are required to wear a Fitbit to seamlessly report data on their steps taken and heart rate to ensure they meet the university’s fitness requirements, which factor into their grades for health and physical education classes.
Researchers at the Massachusetts Institute of Technology (MIT) have developed a microchip called Eyeriss designed to support artificial neural networks, which power artificial intelligence applications, on mobile devices. Neural networks typically rely on graphics processing units (GPUs) to process data, but GPUs designed for mobile devices such as smart phones are typically not well suited for this type of analysis, and applications instead rely on uploading data to the Internet for processing. Eyeriss, which can perform this type of analysis 10 times more efficiently than typical mobile GPUs, would make it possible for mobile devices to locally process the data necessary to support artificial intelligence applications without transmitting additional data.
Researchers at Rensselaer Polytechnic Institute have developed a method to predict which Medicaid patients are likely to make frequent return visits to the emergency room by analyzing the electronic health records (EHRs) of 1.1 million patients. The researchers found that 5.46 percent of emergency room visits were for patients who had previously visited the emergency room within 72 hours, and they were able to identify certain factors indicating an increased likelihood that a patient would return, such as if the visit was related to substance abuse or certain types of infections, and if the visit took place on the weekend or late at night. By identifying patients likely to quickly return for emergency care, hospitals could better manage staffing resources and better educate certain patients about how to manage their care.
The Sensors Exploitation Research Group at the Air Force Institute of Technology have developed a prototype sensor capable of differentiating between human skin and other materials in images and video, which could, for example, help emergency responders more quickly identify people trapped in rubble after an earthquake or help intelligence teams identify humans in surveillance videos. The sensor can detect the amount of water and melanin—which determines skin pigmentation—in different materials, and can even differentiate between individual skin colors, which could help analysts single out particular individuals from a group.
The U.S. House of Representatives has passed a group of bills that include a provision to exempt 60 percent of public companies from requirements to report financial statement information to the Securities and Exchange Commission (SEC) in the XBRL machine readable format. The SEC currently requires public companies to report this information in both XBRL as well as in a text-based format, such as PDF. This dual reporting requirement has frustrated companies and investors alike, who rely on this data to make important business decisions, but eliminating the XBRL reporting requirement rather than the outdated text-based format makes it harder for members of the public and private sectors to analyze this data. The Obama administration plans to veto the package bill if the Senate were to pass it.
An Australian startup called Aipoly has developed an app that uses artificial intelligence image recognition algorithms to help people who are blind identify objects around them with their smartphone’s camera. The app allows users to point their smartphone’s camera at an object and the app will describe what is in the frame over the phone’s speakers in close to real-time. Aipoly is working to expand the app to include translation features and facial recognition.
The National Aeronautics and Space Administration (NASA) has announced that it will launch 13 small research satellites, called CubeSats, along with the launch of the Orion deep space explorer craft to collect data on astronomical weather and deep space radiation, as well as study the Moon and a nearby asteroid. NASA will launch the satellites, which weigh just three pounds each, with Orion to take advantage of its powerful propulsion system so they can access further reaches of space than would otherwise be possible. The Orion launch is scheduled for September 2018.
Google has announced it is piloting a method of combating radicalization online by detecting when users search for extremist material and manipulating its AdWords algorithm to instead display links to anti-radicalization websites. Google normally uses AdWords to determine which sponsored links appear at the top of each search based on search keywords, but now will display links to anti-radical material produced by nongovernmental organizations if a user searches for keywords such as “join ISIS.”
The U.S. Forest Services and Federal Communications Commission announced they will work with the U.S. Geological Survey to expand its ShakeAlert system—an early-warning system for earthquakes that relies on networks of seismic sensors in buildings and the environment—by deploying the sensors in national parks. Researchers at the University of California-Berkeley will also work to integrate smartphone accelerometer data into the ShakeAlert system, and researchers at the California Institute of Technology will work on streamlining the ShakeAlert system as whole.